Ten thousand futures, none of them a promise

Is this an AI prediction?

No, and we will not call it one. It is a deterministic simulation with a fixed seed: the same inputs always produce the same answer, and no model was trained on anything. AI only writes the plain-English summary about the numbers, never the numbers.

How does the simulation work?

It builds 10,000 possible futures by drawing whole calendar years of your own funds’ real total-return history at random and replaying them together, so real crashes, recoveries and the way your funds move with each other all survive.

• That year’s actual inflation rides along, so the return–inflation relationship is preserved
• Total-return prices mean fees and dividends are already inside every number
• Scenarios are percentiles of the 10,000 outcomes, not vibes

Why not just use my funds’ own average return?

Because a roster of modern ETFs may only share history since about 2012 — a stretch with no dot-com bust and no financial crisis in it. Left alone that lucky window would centre a 30-year projection near 12% a year. The middle is instead anchored to forward-looking capital-market assumptions, while the ups and downs still come from real history.

Will two pages ever disagree?

They cannot. The dashboard widget, the risk profile simulator and the scenario comparison all run the same engine with the same seed, so no two screens can quote different futures for the same plan.